2

Is the VA National Formulary Overly Restrictive and Does It Prevent Physicians from Meeting the Health Care Needs of Veterans?

To begin, the committee concluded that the answer to the question posed by Congress and the VA—whether the VA National Formulary was overly restrictive—was dependent, at least in part, on judgment. The restricted budgetary resources of the VHA that make veterans ' health care a zero-sum game in which inflation in one sector obligates deflation in another will, of necessity, influence such a judgment. The IOM committee, as a panel of experts with broad experience, identified and evaluated the various elements, dimensions, or categories of restrictiveness as they are employed by the National Formulary and formulary system and compared them in pharmacy benefits of the VA, private-sector MCOs and PBMs, and two public programs, Medicaid and the Department of Defense. The information in Chapter 3 of this report supports an analysis of the economic factors that underlie some decisions on restriction. The central elements of restrictiveness of the formulary are also susceptible to independent analysis. Although there are deficiencies in the data, the committee found, in many instances, that there was sufficient information to support discussions, analyses and conclusions reached in this chapter.

BACKGROUND

Restrictiveness is a multifactorial attribute of a formulary and formulary system. At its root, the restrictiveness of the VA National Formulary, including the local, VISN, and national lists and systems, and the nonformulary exceptions processes, is a measure of the stringency of the controls on veterans' access to prescribed medicines at the appropriate times. Comparisons with other systems



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary 2 Is the VA National Formulary Overly Restrictive and Does It Prevent Physicians from Meeting the Health Care Needs of Veterans? To begin, the committee concluded that the answer to the question posed by Congress and the VA—whether the VA National Formulary was overly restrictive—was dependent, at least in part, on judgment. The restricted budgetary resources of the VHA that make veterans ' health care a zero-sum game in which inflation in one sector obligates deflation in another will, of necessity, influence such a judgment. The IOM committee, as a panel of experts with broad experience, identified and evaluated the various elements, dimensions, or categories of restrictiveness as they are employed by the National Formulary and formulary system and compared them in pharmacy benefits of the VA, private-sector MCOs and PBMs, and two public programs, Medicaid and the Department of Defense. The information in Chapter 3 of this report supports an analysis of the economic factors that underlie some decisions on restriction. The central elements of restrictiveness of the formulary are also susceptible to independent analysis. Although there are deficiencies in the data, the committee found, in many instances, that there was sufficient information to support discussions, analyses and conclusions reached in this chapter. BACKGROUND Restrictiveness is a multifactorial attribute of a formulary and formulary system. At its root, the restrictiveness of the VA National Formulary, including the local, VISN, and national lists and systems, and the nonformulary exceptions processes, is a measure of the stringency of the controls on veterans' access to prescribed medicines at the appropriate times. Comparisons with other systems

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary can influence judgments about such controls. Nevertheless, if formulary structure or formulary system controls deny or significantly delay access to drugs that, in the reasonable judgment of medical experts, are clinically indicated, then the VA National Formulary meets a definition of “overly” restrictive. Such clinically indicated or medically necessary medicines are not necessarily those identified in television commercials or by pharmaceutical sales representatives, or those preferred by patients or physicians for other reasons. VA National Formulary treatment guidelines and drug class reviews are intended to improve prescribing decisions by caregivers. The evidence that there is room for such improvement is substantial. Physicians are often swayed by industry commercial messages when scientific data would indicate otherwise (Avorn et al., 1982; Avorn and Sounerai, 1983). Prescribers respond to patient pressure that may result from direct-to-consumer advertising (Barents LLC, 1999). Physicians also may provide drugs primarily for a placebo effect or to meet patient expectations for some sort of intervention (Schwartz et al., 1989). Prescribing errors are distressingly frequent throughout the U.S. health care system and have been shown in demonstrations to be correctable through consultation with clinical pharmacists or reengineering of systems of health care (Institute of Medicine, 1999; Leape et al., 1991, 1995, 1999). Elements of Restrictiveness The committee decided that the question of restrictiveness could be approached by examining several characteristics of the National Formulary and formulary system, both on their own merits and in comparison with other formulary systems. The elements of restrictiveness discussed in this chapter include the following: formulary size and coverage of agents in different therapeutic classes; timeliness of addition of new drugs or drug products and reappraisal of formulary listings; and appropriateness and responsiveness of the nonformulary exceptions process and access to nonformulary drugs. Restrictiveness also depends on therapeutic interchange policies and practices that are standardized and protect at-risk groups of patients from drug treatment misadventures. Coverage of nonprescription (over-the-counter [OTC]) drugs and generic substitution are also important. The committee concluded that the key elements in the VA National Formulary are the number of classes closed, number of drugs in closed classes, responsiveness of the nonformulary process, and sensitivity of the therapeutic interchange policy and Elements of Restrictiveness Formulary size Coverage of agents in different therapeutic classes Timeliness of addition of drugs newly approved by the Food and Drug Administration Reappraisal of formulary listings Appropriateness and responsiveness of the nonformulary process Access to nonformulary drugs Sensitivity of therapeutic interchange policy and procedure Over-the-counter coverage, generic substitution

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary procedure to patient risks and prescriber prerogatives. These are central elements in the implementation and management of the National Formulary, and revision of policies and procedures governing them will represent significant changes. Criteria that the committee used for judging the restrictiveness of the attributes of the VA National Formulary include the following: how they compare to those of other organized private and public health care systems; how they compare to reasonableness standards in the literature or in the informed judgment of the committee; how they compare to objective standards where these are available; and how they affect the satisfaction and opinions of patients and prescribing physicians. The committee identified the elements of restrictiveness in the VA National Formulary and formulary system and, for comparison, in private-sector formularies and formulary systems. These are presented in Table 2.1. Private-sector data were collected through a special questionnaire (see Appendix B) circulated in January 2000 by the Academy of Managed Care Pharmacy (AMCP) to six major private-sector PBMs and two small MCOs covering in total about 176 million lives. Care should be exercised in interpreting these data in the sense that covered lives may be overstated due to double counting of two-wage-earner families. Also, some PBMs that provide claims services and not formulary management may have reported their own policies and not the actual client MCO formulary policies. Public-sector (Medicaid) elements are identified and discussed in Chapter 5 and summarized in Table 5.1 of that chapter. In all of these comparisons, clear differences in the involved health care systems in which the formularies are embedded suggest caution in drawing inferences, although the committee has attempted to limit the comparisons to the formularies and formulary systems not the health care systems. Restrictiveness can be approached in another way. The use of restrictiveness elements or the characteristics of formularies and formulary systems that affect the availability of, or ease of obtaining, a drug in a health care system can be categorized, with the more severe limitations being those that absolutely deny access or limit access without medical need-based exceptions. Controls without need-based exceptions, such as absolute limits on numbers of prescriptions in some Medicaid programs, are rare in the private sector and are not part of the VA National Formulary. Box 2.1 lists such formulary limitations on receiving a safe, effective, and medically necessary drug, some with essentially no limit on access, others with complete inaccessibility. The committee concluded that the most important limits incorporated into restrictive designs were exclusions, volume or quantity limits that were unresponsive to medical necessity, drugs not being listed in a closed class, or not included or covered in the formulary, high copayments (these are also not related to medical necessity), and administratively and medically strict prior approval or nonformulary exceptions processes. Although this chapter is not organized by the listing in Box 2.1, an appreciation of these factors— their roles in, and contributions to, the availability or restriction of choice of drugs and drug products, is an important background to the committee's exploration of elements of restrictiveness of the VA National Formulary and comparison formularies. Formularies usually fall into one or more of the listed limitation categories and

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary implement them to varying degrees. The first step in analyzing restrictiveness is understanding these elements. The next is knowing how they contribute variably to different formulary designs that affect the availability of drugs. SIZE AND COVERAGE OF THE NATIONAL FORMULARY The VA National Formulary, dated July 1999, lists about 1,200 items, of which 133 are medical-surgical supplies. The nonsupply items are distributed in 254 classes and subclasses. Some of these classes are vitamins, dentals, vaccines, diagnostics, radiographic contrast agents, and intravenous (IV) or other solutions, that is, items that would not be considered typical pharmaceuticals and are often not included in other formularies. About 15 of the listings indicate that a drug class is under review without referring to any specific agent. Drugs needed in these classes will be found in VISN and local formularies until reviews are completed and national decisions made. About 170 of the items listed are OTC, such as nutritional supplements, vitamins, cough and cold remedies, simple ointments and other topicals, eye and nose drops, antacids and laxatives, and the like. Items such as these may be substituted for more expensive or risky prescription drugs. About half of the items in the VA classes are not separate chemical agents, but represent the same chemical entity in injectable, oral, or topical form. Dosage strengths are generally not identified. Drug dosages that are stocked and immediately available in VHA health care facilities are left to the discretion of each facility's management. TABLE 2.1 Comparison of VA and Private Health Care   VA (%) Private Health Care (% of 176 million lives) P&T committee composition Pharmacists 52a 32 Physicians 44 63 Others 4 5 Excluded DESI 100 10 Experimental 100b 99 Off-label use 0 36 OTC 0 90 Life-style 100c 92 Formularies Closed or partially closed   18 Open-preferred 100d 33 Open-passive   38 No formulary (PA)   13 No formulary (DUR)   1 No formulary (free access)   0 Closed formularies that contain only one drug in the drug class 100 3.5 Drug restrictions (specific prescribers, settings, disease conditions) 100 71 Required use of generic drugs 100 38 Nonformulary Coverage of nonformulary 100e 100 Copay to influence choice 0 100 Cost containment Limits on number of Rxs per patient at any time 0 23 Limits on refills 0 46 Limits on duration of use of some drugs 0 21 Limits on the supply of drugs per Rx 100 71 Presence of a PA process for some drugs 0 53 Waiting period requirements for new FDA-approved drugs 100f 6 Six months or more wait period 100f 1 Active monitoring of new FDA-approved drugs: Drugs for the treatment of AIDS or cancer 100 71 FDA “1P” drugs 0 34 Appeals process Internal appeals process for excluded drugs 100 47 Internal appeals process for nonformulary drugs 100 19 Independent external review of appeals process 0 7 Continuation of care Policy requires continuation of care for a few specific drugs 0g 3 Policy requires continuation of care for all drugs 0g 9 NOTE: DESI = Drug Efficacy Study Implementation; DUR = drug utilization review; FDA = Food and Drug Administration; PA = prior approval; 1P = FDA priority; P&T = pharmacy and therapeutics. aReflects the composition of VISN formulary (P&T) committees. b The VA does not cover experimental drugs but it does not preclude the use of experimental drugs in its research programs. c The VA does not exclude any class of drug, but if a specific agent is not on the formulary, it must be accessed by the nonformulary process. d The VA formulary is a composite of a closed open–preferred, and open–passive formulary. The 100% for veterans should be compared to the summation of these three types. e When medically justified. f In some VISNs, drugs can be placed on the formulary earlier than 1 year. gThe VA does not have a specific policy for continuation, but a nonformulary drug can be continued if approved through the nonformulary process. SOURCE: Private health care data from Academy of Managed Care Pharmacy.

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary This page in the original is blank.

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary Adequacy of the VA National Formulary With respect to the adequacy of the VA National Formulary, the committee asked several questions. Is the overall size of the formulary reasonable? Are the closed classes limited to a reasonably small number for which economic and BOX 2-1 Formulary and Formulary System Limitations or Other Restrictions on Drugs in a Health Care System Drugs are covered and listed in an open formulary or in an open class. Drugs are not listed in the formulary, but the formulary is open. Coverage is assumed, or, at worst, a simple nonformulary or prior approval process is required. Drugs are subject to mandatory generic substitution. Drugs are restricted in some way: to a kind of prescriber or specialist, to certain conditions or diagnoses, to certain services or care settings, under a step protocol requiring a trial of another drug. Drugs are preferred or encouraged in some way. Copayments, or copayments that vary in amount depending on the degree of preference, are required. Drugs are not covered and not listed in a partially restricted or closed formulary; they are available by prior approval or by nonformulary exception. Drugs are members of a closed class and are not listed: they are available by nonformulary exception or therapeutic interchange is required. Drugs are subject to volume restrictions independent of medical necessity; there can be limits on: dispensing frequency, quantity dispensed, dollar value dispensed, number of prescriptions per unit time. Drugs are listed as excluded or no-buy, or are members of excluded or no-buy classes. These are not available except by out-of-pocket, retail purchase by the patient.

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary FIGURE 2.1 The number of closed classes in closed private-sector formularies by percentage of private-sector plans. SOURCE: Academy of Managed Care Pharmacy (2000). therapeutic effects justify the effort (and possible inconvenience to prescribers and risks to patients) of managing that class? Are the numbers of members in a closed class listed on the formulary reasonable and sufficient, particularly in comparison with other formularies? The standard of reasonableness depends in part on the committee's professional judgment and experience with other formularies and in part on information in studies and reports in the scientific literature. The overall size of a formulary is not necessarily a key characteristic. There is no standard that specifies a particular size for a given health care system. Large open formularies may contain from 1,000 to 3,000 drugs and dosage forms (Covington and Thornton, 1995). Those of the Mayo Clinic and the British National Formulary list more than 1,200 and 2,000 items, respectively. In 1997, most MCO formularies (60%) contained less than 1,000 items, and 37% were in the 750–999 category (Hoechst Marion Roussel, 1998). A majority of MCO formularies were closed or partially closed (Novartis, 1999). A more open formulary that includes greater numbers of widely prescribed drugs does not automatically improve clinical practice. In studying Medicaid formularies after OBRA 1990, which traded drug rebates for abolition of restricted formularies, Walser et al. (1996) concluded that only 22% (4/18) of additions of top 200 drugs to state formularies conferred a net therapeutic benefit, that is, led to better patient compliance, were less expensive, or had greater effectiveness, according to panels of practicing physicians. The rest either were questionable or did not offer an additional benefit. The authors could not draw any policy conclusions from these data. Many drugs are widely prescribed because prescribers are influenced by industry sales techniques or for other non-scientific reasons (Avom and Soumerai, 1983; Avorn et al., 1982; Schwartz et al., 1989), or because patient demand is generated by direct-to-consumer adver-

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary tising (Barents LLC, 1999a; www.imshealth.com/html/news_arc/06_07_1999_211.htm). Rucker and Visconti apparently considered that a good-quality hospital formulary listed only about 450 single drugs, but their reports are dated, and in any case apply to a hospital, not a health care system (Rucker, 1982b; Rucker and Visconti, 1978). In 1986, Bakke reviewed the situation internationally and concluded that at that time, about 500 drugs should be available to deliver good care in most advanced countries (Bakke, 1986). This estimate is primarily of historical interest given its age and the fact that it was devised for countries not health care systems. Rhode Island Hospital, the principal teaching affiliate of Brown University, was reported to have 580 medications listed in its formulary in 1986 (Packer et al., 1986). In a review of survey data from 187 large, private, nonprofit, U.S. teaching hospitals, Mannebach et al. (1999) noted that most of their formularies were closed and that P&T committees tried, for the most part successfully, to limit the numbers of drugs listed in therapeutic classes. The VHA allows VISNs (and local facilities) to add items that are not on the National Formulary. Therefore, choices actually available on these formularies (although the same in at least one instance, VISN 14) usually exceed those on the national list by a few to 544 items. VISN formularies frequently include about 10% more items. Precise formulary comparisons are difficult because the VISN formularies list multiple dosage forms and the National Formulary does not. One formulary for which precise figures were provided (VISN 19) contained 108 more drugs and a total of 615 different drugs. Prior to the VA National Formulary, some VISNs were functioning at 70% or less of the present National Formulary size. Most VA local and VISN formularies had fewer listings before the national list was introduced in 1997. Some of these facilities were operating, apparently satisfactorily, with a formulary of 70% or less the size of the present National Formulary (VHA data provided to IOM, 1999). In view of the increase in drugs and drug products available to veterans since the introduction of the National Formulary and in comparison to MCO or hospital formularies, the number of items on the VA National Formulary seemed reasonable to the committee. Closed and Preferred Drug Classes in the VA The VA National Formulary's closed classes, the quality of the listings as affected by the quality of the input of P&T committees, the MAP, and VA PBM (see Chapter 4 of this report), and the access afforded through a nonformulary exception process are probably more important measures of restrictiveness than overall formulary size. Although dozens of therapeutic classes have been or are being reviewed by the VHA, only six classes (four at the present time) have been closed. Two classes are preferred, that is, they are open, but there are national contracts for one or more members of the class.

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary The closed classes, identified in this report's introduction, are (ACEIs), (HMG CoA RIs), (LHRHs), and (PPIs). They comprised about 16% of the prime vendor cost of drugs in 1998, and, because the VHA accounts for the great majority of prime vendor sales, probably about 16% of VHA drug costs in that year. According to GAO, they comprised 13% of drug costs in 1999 (GAO, 1999). * Because there are only about 20 chemical entities in these classes —that is, about 2% potentially (if expected new entries and different routes of administration and dosage forms are included), and less than 1% in reality, of the drugs on the National Formulary—these classes have an economic impact disproportionate to their numbers. This reflects their prices and the importance of their therapeutic effects; the prevalence of the conditions they can treat, especially in the older, almost all-male VA population; and the volume of prescriptions written for them. In fact, nationally, three of these closed classes were in the top five classes accounting for drug expenditure increases and the top seven in class cost in 1998 (Barents LLC, 1999a). According to the last Novartis (1999) survey of MCOs, gastrointestinal medications (such as PPIs), ACEIs, and HMG CoA RIs were in the top five classes by overall cost, number of prescriptions, and per-member per-year expenditures in managed care pharmacy benefits. The closed classes are clearly important, major classes (see also Figure 2.1 for comparison of number of closed classes in PBMs that close classes). Preferred classes are also contributors to drug costs and utilization. As noted, there are two preferred classes (CCBs and alpha blockers), one of which (alpha blockers) was previously closed. The case for creating preferred classes may not be quite as persuasive as that for closed classes. In theory, more local options are provided in these classes because VISNs and facilities are free to supplement the national list. However, economics drive the choice to the nationally contracted drugs in the preferred class and may encourage changes in prescribing by physicians or initiate therapeutic interchanges. A previously closed class, (H2R) blockers, is open since some members are now generic and famotidine, cimetidine, and ranitidine are OTC. Price or cost control in this class is also exercised through contracting, which has a striking effect on utilization (see Figure 4.4, Chapter 4 of this report). National usage criteria that generally depend on therapeutic guidelines or drug class reviews are also issued from time to time by the VA PBM. There were nine separate usage criteria in the fall of 1999 (http://www.dppm.med.va.gov/newsite/criteriadrop.html). These criteria may require certain patterns of prescribing or even therapeutic interchange. They may also dictate the ways in which nonformulary drug products (for example, new anti-inflammatory agents, such as cyclooxygenase-2 [COX-2] inhibitors), can be used acceptably for arthritis and other conditions. In encouraging such behavior, preferred classes, those with national usage criteria, or those with committed-use contracts are similar to closed classes, although they are officially open. * Based on 1998 prime vendor purchase data, the projected fifth closed class (oral 5HT3 RAs) would account for less than 0.5% of the top 200 VA drug costs.

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary After VA drug class review, six classes that were reviewed for possible closure have not been closed, preferred, or subjected to restrictions at the national level, although they might have been. Such classes, which are often managed (closed, restricted, or subject to therapeutic interchange) in other health care systems, include some drugs (selective serotonin reuptake inhibitors [SSRIs], other antidepressants) used in mental health care or expensive antibiotics (for example, fluoroquinolones or cephalosporins) (Achusim, 1992, reviewed in Bootman and Milne, 1996; DeTorres and White, 1984, reviewed in Dunagan and Medoff, 1993; Dzierba et al., 1986; Edwards and Anderson, 1999; Guze, 1996; Kresel et al., 1987, reviewed in Mitchell et al., 1997; Nightingale et al., 1991; Reeder et al., 1997; Stock and Kofoed, 1994; Streja et al., 1999; Zhanel et al., 1989). MCO formularies also often restrict expensive, brand name products through higher copayments (Hoechst Marion Roussel, 1998; Smith, 1993). Drugs in VA closed or preferred classes, although not chemically identical, are similar in therapeutic effect. Price differentials exist among class members. Generics are few or absent, and significant price reductions and aggregate savings through volume commitments are possible (see Chapter 3 of this report). These same classes are also subject to restrictions and therapeutic interchange in other (MCO) systems. Restriction to a few choices through closure or preference in each class is reported to be consistent with good care because class members are thought to be therapeutic alternates. That is, they have similar therapeutic effects based on the available medical evidence (see VHA class reviews and accompanying references at http://www.dppm.med.va.gov/newsite/reviews.html ; see also Briscoe and Dearing, 1996; Gerbrandt and Yedinak, 1996; Hilleman et al., 1997; McMillan, 1996; Moisan et al., 1999; Oh and Franko, 1990; Petitta et al., 1997). Potentially closed or preferred classes that are not economically significant, are low volume, have many generics, or have no comparatively substandard members may not yield economic or quality rewards from being designated as closed or preferred that are sufficient to justify the time, effort, and potential prescriber and patient dissatisfaction involved. Classes for which quality concerns would be raised by a designation as closed or preferred definitely should not be so designated (Carroll, 1999). These classes are not closed or preferred in the VA National Formulary. Restrictiveness of Class Closure The VA National Formulary has some classes with only a single agent, which is unusual in other formularies (see Table 2.1 and discussion of some MCO formularies and Medicaid in Chapter 5 of this report). These may be classes with only one or two members, such as chloramphenicol or PPIs (at the time of original listing). If the classes are open, facilities and VISNs are free to expand the listings. Listing only a single drug in a class is not prima facie evidence of excessive restriction, but it tests the ability of the nonformulary process to provide access to therapeutic alternates if medically needed. The absence of

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary formulary alternatives (if such drugs exist) deserves examination. It is unlikely that a single member of a multidrug class will treat all patients for whom the class is indicated without exception (McAllister et al., 1999), and some investigators have reported that restriction of choice to a single agent compromises care and raises costs (Streja et al., 1999). Therefore, a smoothly functioning non-formulary process is important to preserving quality of care in this situation. The listing of only one agent in closed classes, such as PPIs and LHRHs, results in almost all veterans being on the formulary agent lansoprazole or goserelin (see Figure 4.2 and Figure 4.3 in this report). All four potential choices (as of July 1999), including nonformulary omeprazole and leuprolide, are on the 200 top dollar expenditure list, however, so nonformulary costs are at least measurable. In general, VA nonadherence reports identify about 5% nonformulary dispensing in closed classes with a single formulary choice. Decisions on PPIs and LHRHs were made on the basis of good-quality drug class reviews (see Chapter 4 for analysis; for actual reviews, see http://www.dppm.med.va.gov/newsite/reviews.html). There is also a specific VA cost-effectiveness study that compares PPIs (Vivian et al., 1999). The VA nonformulary exceptions process is discussed briefly below and in Chapter 4 of this report. In a lengthy review of restrictive formularies, many of which are also reviewed in Chapter 5 of this report, Levy and Cocks (1999) claimed that the VA National Formulary severely limits choice among brand name agents in six drug classes, three of them current VA National Formulary closed classes. These authors provided no evidence of effects on veterans' health outcomes by National Formulary restrictions in these classes. Although Levy and Cocks (1999) raised some interesting concerns, some of which are discussed further below, the committee concluded that they had not made a persuasive case for meaningful and severe limitation of choice in the six cited classes. Median coverage by 13 Medicaid programs was 27 of the 27 brand name products in the six classes, and by the National Formulary, 8 of 27. It is not clear that these Medicaid programs are representative, and in any case, states must include in their Medicaid formularies all drugs that manufacturers list in the Federal Supply Schedule. Of four cited National Formulary classes with only one representative, three are either open or preferred so that VISNs and facilities are free to add drugs that are used and preferred locally. The other class, PPIs, which is closed, had only two brand name products (additional products have since been marketed) from which to choose, and they are considered therapeutic alternates (Chon and Suzuki, 1998; Vivian et al., 1999; see also the VHA drug class review and its accompanying references at http://www.dppm.med.va.gov/newsite/reviews.html). A VA drug class review recommended one nitroglycerin patch in that open class, based on cost and patient preference since available products are similar and, with the exception of a few high-dose patches, are rated bioequivalent by the Food and Drug Administration (FDA, 1999a). The other two classes (alpha blockers and H2R blockers) are currently represented by two (prazosin, terazosin) and three members (cimetidine, famotidine, ranitidine), respectively.

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary cally unjustified nonformulary use. Computerized data on drug utilization by prescriber could make this possible. This recommendation should not imply committee disapproval of reasonable formulary enforcement. Price negotiations will not be effective if prescribing is not directed to appropriately selected formulary drugs or drug products. THERAPEUTIC INTERCHANGE In managing the National Formulary and the pharmacy benefit, the VHA (with the National Acquisition Center) closes classes, and/or negotiates national standardized (committed-use) contracts or blanket purchase agreements with manufacturers for drugs in the classes or other classes, and/or issues national usage criteria or restrictions. In those situations, anticipated or promised volume supports better prices, which is, after all, one of the main rationales for the formulary. To achieve the promised volume (or market share) and realize the better price and, therefore, the savings, there is an expectation that VHA prescribers will discontinue using nonformulary or noncontracted drugs so that veterans are not started on these agents or that they may convert veterans from nonformulary or noncontracted drugs to the preferred or formulary agents. This expectation is no different than the expectation and practice in many U.S. hospitals and MCOs, as noted elsewhere. Interchange without individual prescriber permission at the time of dispensing is not uncommon in the VHA, however. Exchanges can be made by the pharmacy on authorization by the chair of the P&T committee without consulting each prescriber. The (prescribing and) dispensing by the pharmacist of a “therapeutically equivalent, ” pharmaceutically different drug than that prescribed for a patient by the physician (that is, therapeutic interchange with therapeutic alternates) is specifically authorized by the VA (VA Manual, M-2, Part 1, Chapter 3, Clinical Programs, Pharmacy and Therapeutics Committee, December 13, 1993). Most therapeutic interchange in the VHA has been in closed classes or with a few popular drug products in other classes. These include situations in which there are price differentials, high volume, national contracts or usage criteria, current or former class closures, or VISN as well as national purchase agreements or contracts. In these situations the local facility, VISN, or VHA is anticipating target volume and target contract prices or responding to significant existing price differences to make savings in drug budgets (see reference to VHA reports below). VHA investigators have published a number of reports describing therapeutic interchange programs in VA facilities in the 1990s. Additional reports have been presented at meetings but not published, and some less formal program data are also available. These reports often suffer from one or more of the following: small numbers, short follow-up, incomplete data and monitoring, and lack of controls, among others (Bartlett et al., 1996; Boston and Collins, 1995; Brunsting and Johnson, 1997; Cantrell et al., 1999; Desai et al., 1997; Edwards et al., 1998; Ganz and Saksa, 1997; Gray, 1999; Gustin et al., 1996; Ito et al.,

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary 1999; Jones, 1999; Kellick et al., 1995; Kinnon et al., 1999; Lederle and Rogers, 1990; Lin et al., 1999; Lindgren-Furmaga et al., 1991; Minnich et al., 1997; Patel et al., 1999; Rindone and Arriola, 1998; Sprague and Gray, 1998; Stanaszek et al., 1997; Stock and Kofoed, 1994, Vivian et al., 1999). Therapeutic interchange in these reports, and in the VA in general, has been driven primarily by cost considerations (through contract and usage criteria adherence). Some VA interchanges have been clinically successful (for example, see Boston and Collins, 1995; Cantrell et al., 1999; Ganz and Saksa, 1997; Jones, 1999; Kellick et al., 1995; Kinnon et al., 1999; Lin et al., 1999; Patel et al., 1999; Rindone and Arriola, 1998; Sprague and Gray, 1998: Vivian et al., 1999), but some have had clinical problems (for example, see Bartlett et al., 1996; Brunsting and Johnson, 1997; Desai et al., 1997; Minnich et al., 1997, Stanaszek et al., 1997; Stock and Kofoed, 1994). Economic outcomes have also been mixed because it may take time for the savings from a less expensive drug to pay back switching costs (Lindren-Furmaga et al., 1991), or savings from drug costs may have been offset by increases in other costs (Bartlett et al., 1996). On balance, however, most reports of therapeutic interchange have been encouraging, or at least reassuring, especially those describing more recent efforts. A survey of 192 HMOs reported in 1988 that 30.5% used therapeutic interchange. Those that did not were concerned about physician dissatisfaction, interference with physicians' prerogatives to prescribe according to their best judgment, and the legality of substituting (Doering et al., 1988). Other surveys find that therapeutic interchange is an accepted practice in managed care that is becoming more prevalent each year. The Hoechst Marion Roussel survey (1998) reported 35.2% of plans allowing therapeutic interchange in 1997. Novartis (1998) reported that 47.7% of HMOs made use of therapeutic interchange in 1997; this survey projected an increase to 61.4% for 1999, and reported 56.5% of plans using therapeutic interchange in 1998 (Novartis, 1999). A survey of pharmacy benefits managers of employer plans in 1997 found 53.5% using therapeutic interchange (Wyeth-Ayerst, 1998). The IOM committee noted that some of the VA drug classes involved in therapeutic interchange are also among those most often (more than 80%) subject to interchange in MCOs that practice interchange. These are antiulcerative (PPIs and H2R blockers), antihypertensive (ACEIs, CCBs, and alpha blockers), and cholesterol-lowering drugs (HMG CoA RIs) (Hoechst Marion Roussel, 1998; Wyeth-Ayerst, 1998). Some classes that are commonly controlled through interchange in managed care, such as expensive antibiotics, are left to local discretion by the VHA. Patterns of use also reflect local microbial resistance patterns. Other commonly interchanged classes in MCOs, such as antihistamines and anti-inflammatories, are limited to a few members in the VA National Formulary but are open to respond to local preferences at the VISN or VA medical center level (Hoechst Marion Roussel, 1998). Many of these classes are also reported involved in interchange in hospital programs. The committee concludes that they are also classes that have therapeutic alternates and price differentials, where therapeutic interchange is both medically and economically reasonable,

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary provided there is a provision for access to nonformulary therapeutic alternates when clinically necessary. Therapeutic interchange has been carried out in hospitals for decades, and reports from hospital and other programs have been mostly supportive, reflect-ing experiences similar to those described by the VHA (reviewed in Achusim, 1992; Brown and Clarke, 1992; Bull et al., 1999; Green et al., 1989; Guastella, 1988; McAllister et al., 1999; Rich, 1989; Smith et al., 1989; Wall and Abel, 1996). Problems are pointed out in some of these reviews and elsewhere (for example, Barksdale AFB, 1998; Richton-Hewitt et al., 1988). Recent surveys, some of them quite extensive, find that therapeutic interchange is allowed or practiced in almost all U.S. hospitals (Nash et al., 1993 [76%, about two-thirds without notifying the prescriber]; Reeder et al., 1997 [74.5%]; Sloan et al., 1997 [79%]), and Mannebach et al. [1999] found that 69% of hospitals had formal therapeutic interchange policies). Therapeutic interchange is supported by professional groups, provided the permission of the prescriber is obtained. As discussed earlier in this report, in hospitals and other well-controlled settings, such as some staff or group model HMOs, P&T committees in communication with the medical staff can design therapeutic interchanges. These interchanges are often implemented after ad-vance approval by the P&T committee and medical staff, or they can be specifi-cally approved by individual prescribers. In other health care systems, outpatient and less well controlled settings, PBMs, IPAs, and the like, the prescribers per-mission is almost always sought at the moment of interchange (ACCP, 1993; ACP, 1990; AHA, 1974; AMA, 1994; AMCP, 1997; ASHP, 1982; Lipton et al., 1999; Zellmer, 1994). State laws on interchange vary in details to some extent. Washington is the only state that has specifically recognized therapeutic interchange, however, and has limited it to hospitals. States are consistent in requiring prescriber approval of therapeutic interchange. In controlled settings, this may be achieved according to protocols designed under collaborative practice agreements or other arrangements in which physicians are advised and approve (AMA, 1994). In hospitals, for ex-ample, physicians agree to abide by hospital policies and procedures when join-ing the hospital staff (Fink et al., 1998). MCO, PBM, and Medicaid providers must abide by state laws, but in a federal system like the VA, VHA policies on therapeutic interchanges (see below) preempt these state laws. In theory, therefore, therapeutic interchange in the VHA does not appear re-strictive in comparison to other health care systems. Interchange is a common practice. It is used in relatively few therapeutic classes in the VHA, and these classes are often subject to interchange in hospitals and MCOs and are medically defensible. Articles in the medical literature from the VHA or elsewhere do not report serious problems with most interchanges. On the other hand, there are no national VHA guidelines on therapeutic interchange, and the committee did not find any written policies at the VISN level. Although interchanges may be initi-ated and specified at either the national, the VISN, or the local level, they are designed and implemented at the local facility level. As such, they respond to

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary local practices and need not be nor are they, consistent across the VHA. Some veterans subject to interchanges, when surveyed, report not having received adequate, or any, information on the replacement drug (see Chapter 4 ; W.N. Jones, personal communication, VISN 18, 1999). This suggests that informing veterans of interchanges could be improved even in VISNs that have experienced and thoughtful pharmacy leadership that tries to inform patients. Consideration might be given to having the responsible physician deliver the information in person. Variation in some elements of therapeutic interchange to reflect local preferences and practice patterns may be desirable. Consistency in other elements might be important in ensuring program quality and patient and prescriber acceptance. In Chapter 1 of this report it was suggested that these elements might profitably include adequate advance notice and education on relative dosing and other factors for physicians and patients. Attention to patient compliance, provision for exceptions to conversion, and nonformulary access to an alternate or return to the original drug were also important considerations. Other elements included protections for at-risk patients, avoidance of frequent interchange or switching sick patients who were stable on a particular drug, and various methods of clinical and economic monitoring. The interchangeability of drugs in a class has to be evaluated with some care and expertise (McAllister et al., 1999). VA contracts are annually renewable. Drug prices, and therefore contracts, may change. At some point, the VHA will have to evaluate how frequently veterans taking a drug chronically should be subjected to interchange or the total number of interchanges that is reasonable in an individual patient. It is always important that interchanges may affect a stable drug treatment regimen, and from the veteran 's perspective are involuntary and may not be well understood. Lacking national or regional guidance or convincing evidence of a flexible and responsive nonformulary process, therapeutic interchange programs in facilities may appear inconsistent or restrictive to prescribers and patients regardless of where they originate. Concerns about clinical monitoring, compliance, economic data, assessment of patient satisfaction, dosing problems, and varying observation periods were expressed in some published or unpublished reports of VHA therapeutic interchange (Bartlett et al., 1997; Brunsting and Johnson, 1997; W.N. Jones; personal communication, VISN 18, 1999; Lederle and Rogers, 1990; Minnich et al., 1997; Rindone and Arriola, 1998; Vivian et al., 1999). For the most part, they were of minor import, and the overall conclusions of most of these reports were reassuring. Patient and physician complaints about the National Formulary and access to needed, or at least desired, drugs are often related to therapeutic interchange programs and indicate a level of dissatisfaction. Patient complaints about access to drugs are a very small fraction (0.4%) of veteran complaints to patient advocates. Physician surveys tend to reinforce the concern that some therapeutic interchange programs may appear, or in fact be, restrictive from the perspective of prescribers. Interchange without individual prescriber permission may lead to patient or prescriber dissatisfaction unless the setting is controlled and those

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary involved and affected have been adequately consulted and informed. These issues are discussed in Chapter 4 of this report and reviewed briefly below. Much of the detail on therapeutic interchange programs and the nonformulary system is at the local level. Analysis of the restrictiveness of the National Formulary in comparison to other formularies and assessment of the contribution of therapeutic interchange to restrictiveness would benefit from data from the majority of the 172 VA hospitals. Time and resources did not allow the IOM to gather these data. Nevertheless, the committee found that some conclusions on this important issue were possible. The success of the National Formulary in providing leverage to negotiate low drug prices depends on the ability of the VHA to make choices among drug products and deliver volume, or drive market share as it is often described. Conversion of prescribing to the formulary or contract agent and actual interchange of drugs in existing treatment regimens support the responsiveness of the health care system to such choices and contractual terms. Restrictiveness will depend on how well the conversions are made. Factors include adequate access to original or other nonformulary agents when indicated, protection of at-risk patients who may not do well in interchange programs, and education of prescribers and patients among others. Concerns in some of these cases have been raised in the discussions above, and improvements suggested. On balance, however, the committee concluded that available information did not provide any convincing evidence that access to needed drugs is overly restricted. OTC DRUG COVERAGE, GENERIC SUBSTITUTION, AND PHYSICIAN SATISFACTION As noted earlier, there are 170 OTC listings on the VA National Formulary. These include nutritional supplements, ointments and lotions, nose drops and sprays, vitamins and minerals, laxatives and antacids, insulin and diabetic supplies, eye and ear drops and irrigating syringes, cough and cold remedies, analgesics and antipyretics, topical antibacterials, and antihistamines among others. The National Formulary also includes three H2R blockers and nonsteroidal anti-inflammatory drugs (NSAIDs), but as the prescription not the OTC agents. There may be no price advantage to the OTC version for these drugs, given that they are mostly generic, and the OTC version is generally half the usually prescribed dose. Coverage of OTC drugs is categorized as both a pharmacy benefit and a medical benefit in surveys of MCOs. Few HMOs offer OTC drugs as a specific pharmacy benefit, 7.3% in 1997 (Hoechst Marion Roussel, 1998) or 9.9% in 1999 (estimated) (Novartis, 1998). Nevertheless, OTC drugs in a number of therapeutic classes are often available in MCOs. The estimated percentage of HMOs including coverage in 1998 for insulin and diabetic supplies was 90.9%; for antihistamines, 50%; for H2R blockers, 36.4%; for cough and cold remedies, 40.9%, for NSAIDs, 31.8%; and for vitamins, 22.7% (Novartis, 1998). In 1997, only 11% of employer pharmacy benefit plans excluded insulin and syringes,

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary and 30% excluded prenatal vitamins (Wyeth-Ayerst, 1998). PBMs infrequently manage OTC benefits; when they do so, these are primarily insulin and diabetic supplies, but occasionally also cough and cold remedies and rarely H2R blockers and antihistamines (Novartis, 1998). Recent data from a subset of PBMs and MCOs surveyed by the AMCP are consistent with low levels of PBM OTC coverage (Table 2.1). About 12 state Medicaid programs cover very few or no OTC drugs. Most other states cover a majority or occasionally all of the listed categories of OTC drugs, sometimes with prior authorization, sometimes in institutions only. Usually only a limited number of drugs in a category are covered. Medicaid recipients in managed care would presumably receive the same OTC benefit package as those in fee-for-service Medicaid. Categories of OTC drugs variably covered by states include allergy drugs, cough and cold preparations, analgesics, antacids and H2R blockers, topicals, feminine products, laxatives, vitamins, and a few others (NPC, 1997, 1998). The committee concluded that the VA National Formulary appears to be less restrictive in covering OTC drugs than MCOs or the Medicaid program. Generic Substitution The VA National Formulary uses generic terminology and practices generic substitution. Brand name drugs are used when there is no generic alternative, as nonformulary exceptions, and in five instances where there are concerns about the bioequivalence of generic versions of the drug (although some of these are under review). Standardized national contracts or local purchase agreements are available for some generics when good prices can be negotiated from a single generic manufacturer. This process is simply another form of generic substitution and, therefore, unexceptional. Generic prescribing and generic substitution are reasonable from a medical, pharmacological, and economic standpoint. They constitute a near-universal practice, sanctioned in state law and by the FDA (FDA, 1999a). The use of generics in the VHA exceeds that in the private sector or health care generally. The VHA in the four quarters ending in July 1999 purchased from the prime vendor 64% generics and 36% branded products amounting to 9% of total expenditures for generics and 91% for branded product (VHA data submitted in response to IOM questions). Comparable U.S. figures, cited earlier, were 47% of volume and 8% of dollar sales (Barents LLC, 1999a). Generic substitution was described as an accepted practice in the introduction to this report. In hospitals, it is usually automatic, that is, carried out on dispensing regardless of the specification on the prescription (for example, 94% of hospitals in the survey reported by Sloan et al., 1997). About half of the states require dispensing of the genetic drug, if there is one, in their Medicaid programs (NPC, 1998), and all states have laws addressing this issue in varying ways. In 1997, 97.7% of MCOs made use of generic substitution, and 43.8% of prescriptions were for a generic drug (Novartis, 1998). In 1998, HMO generic product utilization was 45.6%, accounting for 21.9% of drug expenditures (No

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary FIGURE 2.6 Average prices per prescription are lower for the generic version. SOURCE: Generic Pharmaceutical Industry Association. vartis, 1999). All PBMs have genetic substitution programs, and 50% of employer plans use these (Novartis, 1998). Mandatory substitution was reported for 38% of the covered lives in the recent AMCP survey (Table 2.1). Hoechst Marion Roussel (1998) reported that 63.8% of HMOs required genetic substitution in 1997 and 43.8% of prescriptions were filled generically. The committee concluded that generic prescribing and substitution have no implications for the restrictiveness of the National Formulary. Cost savings are undoubtedly realized (Smith, 1993), given the known differences between average prescription prices for generic and brand name drugs (see Figure 2.6). Veterans' Complaints Complaints from veteran patients collected by the VHA patient advocates and the VSOs are reviewed in Chapter 4 of this report. Patient complaints are anecdotal self-reports. Lacking a denominator (that is, a defined group at risk) except for the total population of users of VHA facilities, they are numerator data only. As collections of incidents, representing an unknown percentage of an unknown relevant population, they can function only as generators of hypotheses about the subject matter of complaints, not as scientifically valid evidence. To become evidence, data on events described in complaints would have to be gathered in careful surveys of scientifically selected representative cohorts, validated by chart review or patient examination, and analyzed statistically in comparison to a control group. In any case, veteran complaints about access to drugs comprise only about 0.4% of total complaints and overwhelmingly concern nonformulary drugs, the largest category (about 10% of drug-related complaints) being about access to sildenafil. These data suggest that some veterans believe they are not getting the drugs they need (or want) or that their physicians think they need and are prescribing. Complaints also suggest that veterans feel inconvenienced.

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary The committee concluded that these data were consistent with taking a closer look, as noted earlier, at some aspects of the formulary process including addition of new existing and new FDA-approved drugs, the nonformulary process, and programs of therapeutic interchange. Action to change or improve these elements of the national formulary system would then be based on such analyses, not on complaints or anecdotes. However, if the numbers of complaints are any guide, these issues are not major concerns of veterans at present. The committee recognizes that as a practical matter, the VHA cannot routinely investigate issues based on complaints that amount to less than 1% of total complaints. These issues merit the attention of the VHA for the reasons cited in this IOM report. They are also concerns of a significant minority of VA physicians, have political salience and wider health care implications, and are part of a new clinically and economically important process, the National Formulary, that should be evaluated like any significant change. VA Physician Satisfaction In 1998, 1 year after implementation of the National Formulary, the VA PBM in consultation with the RAND Survey Research Group surveyed 4,536 VA physicians, more than half of them general internists, and received responses from 2,952, or 45.2% (Glassman et al., 1999). The objective of this survey was to evaluate the effect of the National Formulary on access to drugs, quality of care, and physician efficiency as seen from the perspective of VA physicians caring for veterans in the VHA. Although most VA physicians (82%) were aware of the formulary, only 32% had actually referred to it, according to the survey. Most clinicians agreed that they could prescribe the drugs needed for their patients, but 29% felt that the National Formulary impinged on quality of care and 37% felt it was more restrictive than formularies in the private sector. Overall, most VA physicians were either neutral or positive about the effects of the VA National Formulary on care to veterans. Somewhat similar results emerged from a much smaller January 1999 telephone survey. Yankelovich Partners (1999) completed interviews of 418 VA physicians from a sample of 2,810 who could be reached or agreed to respond. The total available cohort consisted of a privately generated roster of 6,288 physicians in a wide variety of specialties. This survey, in part designed by and wholly financially supported by Pfizer, Inc., found fewer VA physicians very familiar with, and more with negative experiences or perceptions regarding, the National Formulary (Yankelovich Partners, 1999). The committee noted problems with the RAND study, some of which had also been identified by the VHA. The sample was drawn from subscribers to a VA journal, and it was not clear whether this was, therefore, a random sample or whether it was a group of physicians especially interested in VA matters. Less than half of those contacted responded, which represents an insufficient response rate. Respondents were allowed to report whether the formulary improves care or access, not whether it reduces quality or access, a serious flaw. Minority, but still

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary substantial, percentages of respondents reported they could not prescribe needed drugs (29%) and that their workload had increased (34%). This seemed a high number and was of concern to the committee. Some VA physicians might have confused the national list and system with VISN or local formularies and systems. For the purposes of this report, as defined by the terms of the VA agreement with the IOM, all formularies are considered part of the National Formulary, but identification of the exact locus of problems would be important in considering solutions. In any event, the committee concluded that this survey supported taking a closer look at certain aspects of the National Formulary, among them the nonformulary process, therapeutic interchange programs, and the relationship between VISN or local activities and National Formulary policies and procedures. Similar conclusions were reached by the authors of this survey report (Glassman et al., 1999). The committee noted that the Yankelovich study was also unpublished and not peer reviewed. It suffered from small numbers. Only 14.8% of those who were reached ultimately completed the interview, an inadequate sample. This study asked respondents about negative effects and so could be expected to have results less supportive of the National Formulary and formulary system (Yankelovich Partners, 1999). Truly credible surveys of VA physicians await the design and implementation of such instruments by completely independent experts in a process independent of the VA or any other parties at interest. A survey currently being carried out by GAO for the second report on the VA National Formulary could satisfy this need. A few additional surveys of physician satisfaction were available to the committee. Others have observed that if prescribers report that a formulary does not impair quality of care, there may be less reason for concern about a formulary and formulary system (Gross, 1998). Furthermore, it is generally agreed, and the committee concurs, that “physician support is essential to the success of a formulary, and refusal by doctors to cooperate with formularies may greatly reduce anticipated savings” (Nash et al., 1992). In 1990, Nash et al. (1992) sent a short, 20-question survey to 5,000 primary care physicians involved in managed care and selected randomly from a list of 51,000 national MCO physicians. Only 9.96% of those contacted returned usable surveys. This survey found high rates (49%) of negative opinions and concern that formularies address costs only, although most respondents reported that formulary compliance required little effort. A 1997 report described a survey of New York physicians' opinions on managed care formularies also with a very low (17%) response rate (Green, 1997). This and an accompanying pharmacist survey found substantial dissatisfaction with formulary choices and therapeutic interchange. The low response rate in these two surveys prevents drawing any reliable conclusions, although the results may help to identify areas that are in need of further examination. Black et al. (1988) carried out an informal survey of opinions of physicians in an IPA about a newly introduced formulary with more than 1,300 choices. Among 179 (35.8%) of 500 physicians surveyed, 49.7% found the formulary

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary somewhat or very difficult to work with, 59.8% found it somewhat disruptive of prescribing practices, and 24% indicated that the quality of patient care was reduced or (1.1%) greatly reduced. Many physicians reported that the formulary was too restrictive in some (39.7%) or all (3.9%) areas. These physician surveys suggest that physician acceptance of formularies is an area that warrants continued attention. It is generally accepted that patient (and likely provider) satisfaction is an indicator of quality of care. A national survey of consumers found that they usually did not perceive their physicians' prescribing to be limited by prescribing restrictions. Only 15.6% overall, and 21.7% in managed care, knew of some restrictions (Novartis, 1998). However, consumers also rated prescription drug coverage as one of the top three reasons for selecting an HMO, and what they liked best about the pharmacy benefit was low prices, that all their prescriptions were filled, and that they received the drugs they wanted (Copyright © CareData Reports, Inc., in Wilson and Burke, 1999). Surveys like this and experience from the VHA, in times of rapidly escalating drug costs, indicate that the pharmacy benefit is important to enrollees in managed care and veterans using VA health care (Novartis, 1998). It is clear that there is a direct relationship between restrictions on how services are provided and received and provider and patient satisfaction (Navarro and Cahill, 1999). The committee appreciates this normal human tendency to want unfettered access to all available benefits without economic or other restriction. Nevertheless, VA patient complaints and VA (and other) prescriber dissatisfaction should be taken seriously. They suggest that some elements of formulary design and implementation in health care generally and in the VHA specifically should at least be reviewed. SUMMARY As is almost always the case in public policy, data to inform and support conclusions and recommendations about the National Formulary are incomplete. Nevertheless, the committee concludes that the information in this chapter and other chapters of this report is sufficient to suggest directions and could be usefully considered by the VHA and congressional policy makers in evaluating possible decisions about the VA National Formulary and formulary system. All of the questions that Congress asked the IOM committee to consider are highly relevant and important. This chapter on restrictiveness defines and addresses one key overall effect of the VA National Formulary. This was first among the effects or issues raised in House Report 105-610 and the VA–IOM contract. In discussions here, the committee has described how formularies, or lists of drugs, can be managed through the use of elements of restrictiveness and other management approaches, to affect access to and cost of a pharmacy benefit. The restrictiveness of the formulary can be assessed, admittedly to some extent as a judgment, but also by objectively evaluating what controls are employed, how they are used, and how they compare to restrictions and controls used in man

OCR for page 37
DESCRIPTION AND ANALYSIS OF THE VA National Formulary aging other formularies, public and private. In analyzing the National Formulary and formulary system in this way, the committee found that it had characteristics that were sometimes more but often less restrictive than other formularies and formulary systems. The committee has used the information and analyses here to support its conclusions and recommendations. They can be found in the Executive Summary of this report, beginning with the overall conclusion that the National Formulary does not meet the committee's definition of overly restrictive. The committee hopes that the information, analyses, conclusions, and recommendations here might also guide policy evaluations by others.